INDEX

Explanations

proper nouns or capitalized phrases unrelated to each other

oai_token-act-pair · gpt-3.5-turbo

words related to upcoming events or actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 expulsion

-0.70

 reciproc

-0.68

 srfAttach

-0.65

hement

-0.64

 shedding

-0.63

 shock

-0.63

 defamation

-0.63

 suppressed

-0.62

imposed

-0.60

 instantaneous

-0.59

POSITIVE LOGITS

Own

0.91

Else

0.91

tons

0.81

!,

0.72

angered

0.72

 Unknown

0.71

icia

0.71

 Dying

0.71

Dead

0.70

 Happ

0.70

Activations Density 0.284%

proper nouns or capitalized phrases unrelated to each other

words related to upcoming events or actions

No Comments

No Known Activations