INDEX

Explanations

mentions of actions or circumstances that involve harm, crime, or emotional distress

oai_token-act-pair · gpt-3.5-turbo

concepts related to identity and experiences of marginalization

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 awarding

-0.69

 ranking

-0.68

rg

-0.66

 issuer

-0.64

 testament

-0.62

 authorizing

-0.61

-0.60

Wen

-0.59

POSITIVE LOGITS

ictionary

0.91

ãĤ¶

0.72

è£ħ

0.69

 Yourself

0.69

cribed

0.67

quished

0.67

ãĥīãĥ©

0.67

omorphic

0.67

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

0.66

thood

0.66

Activations Density 0.269%

mentions of actions or circumstances that involve harm, crime, or emotional distress

concepts related to identity and experiences of marginalization

No Comments

No Known Activations