INDEX

Explanations

recommendations and summaries presented in a list format

oai_token-act-pair · gpt-3.5-turbo

sections or bullet points within a document that introduce lists or recommendations

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

 tremend

-0.75

oshi

-0.75

izons

-0.73

odox

-0.69

 inward

-0.69

alysed

-0.69

oud

-0.68

amic

-0.67

 opponent

-0.65

ificantly

-0.65

POSITIVE LOGITS

 Firstly

0.89

 namely

0.85

 Whereas

0.79

SHIP

0.79

 Provided

0.78

 Logged

0.77

 âĹı

0.77

 Essentially

0.76

↵Âł

0.76

If

0.75

Activations Density 0.097%

recommendations and summaries presented in a list format

sections or bullet points within a document that introduce lists or recommendations

No Comments

No Known Activations