INDEX

Explanations

words related to legal matters, policies, and specific actions or events

oai_token-act-pair · gpt-3.5-turbo

specific nouns and references related to current events or legal contexts

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

VIDIA

-0.57

taboola

-0.49

ģĸ

-0.47

SAM

-0.45

 =================================

-0.45

 confir

-0.44

Interstitial

-0.43

Vert

-0.43

 Roche

-0.42

Ble

-0.42

POSITIVE LOGITS

 badge

0.58

 onto

0.52

owment

0.51

iddle

0.49

izont

0.49

into

0.47

 reins

0.46

 commandments

0.46

 salute

0.46

wikipedia

0.45

Activations Density 1.662%

words related to legal matters, policies, and specific actions or events

specific nouns and references related to current events or legal contexts

No Comments

No Known Activations