INDEX

Explanations

references to different types of content

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Content" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Mead

-0.65

wcs

-0.64

 Siem

-0.62

umph

-0.62

JD

-0.61

rolet

-0.61

 damn

-0.60

KK

-0.60

roe

-0.60

ammad

-0.59

POSITIVE LOGITS

edly

1.40

ions

1.17

 creators

1.15

ment

0.99

ious

0.99

ibilities

0.90

iously

0.89

 creator

0.89

 providers

0.87

 creation

0.85

Activations Density 0.041%

references to different types of content

occurrences of the word "Content" in various contexts

No Comments

No Known Activations