INDEX

Explanations

the word "don't"

oai_token-act-pair · gpt-3.5-turbo

the presence of the contraction "don't."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

itiz

-0.70

 Penguin

-0.66

 Presence

-0.66

 spirited

-0.63

 Reloaded

-0.61

 Alleg

-0.61

 Sparrow

-0.60

Reviewer

-0.59

 Laun

-0.58

çĦ

-0.58

POSITIVE LOGITS

 necessarily

1.19

 bother

1.07

 know

1.04

 hesitate

1.01

 discriminate

0.99

 deserve

0.98

 appreciate

0.96

 seem

0.95

intend

0.93

 belong

0.93

Activations Density 0.097%

the word "don't"

the presence of the contraction "don't."

No Comments

No Known Activations