INDEX

Explanations

The neuron activates strongly for the sequence "ign" which appears in words like "Mignon," "vignette," "design," "undignified," and similar terms.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @trainjumperx

phrases ending in "-ign" or "-ized."

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ign" or related variations, indicating a focus on issues related to neglect or disregard

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Mighty

-0.75

cham

-0.74

HAHAHAHA

-0.74

 Springfield

-0.71

bian

-0.71

 Span

-0.65

XL

-0.64

 Akin

-0.61

RC

-0.61

PLA

-0.60

POSITIVE LOGITS

ments

1.21

ificant

1.17

antly

0.99

eous

0.99

atures

0.96

atories

0.96

mentation

0.95

entials

0.95

ame

0.94

ified

0.92

Activations Density 0.017%

The neuron activates strongly for the sequence "ign" which appears in words like "Mignon," "vignette," "design," "undignified," and similar terms.

phrases ending in "-ign" or "-ized."

instances of the word "ign" or related variations, indicating a focus on issues related to neglect or disregard

No Comments

No Known Activations