INDEX

Explanations

words related to actions or processes carried out in a systematic manner

oai_token-act-pair · gpt-3.5-turbo

references to methods or approaches described in a systematic way

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

addons

-0.75

arus

-0.72

verbs

-0.70

raltar

-0.70

algia

-0.69

apego

-0.69

perty

-0.67

orters

-0.66

ãĥĥãĥĪ

-0.64

 Regions

-0.62

POSITIVE LOGITS

 thereafter

0.79

 throughout

0.76

 resembling

0.75

 ILCS

0.72

 whatsoever

0.66

 concurrent

0.64

 affecting

0.63

 across

0.63

 during

0.63

Activations Density 0.241%

words related to actions or processes carried out in a systematic manner

references to methods or approaches described in a systematic way

No Comments

No Known Activations