INDEX

Explanations

phrases related to group membership or identity

oai_token-act-pair · gpt-3.5-turbo

terms related to social roles, statuses, and classifications of individuals or groups

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

nia

-0.88

ntax

-0.73

asus

-0.73

 Temper

-0.69

BUG

-0.65

osi

-0.62

Demon

-0.61

Tenn

-0.60

Austral

-0.59

Synopsis

-0.59

POSITIVE LOGITS

ensitive

0.87

hips

0.85

uggest

0.85

ometimes

0.83

 themselves

0.83

 extraord

0.81

hip

0.79

chool

0.78

pring

0.77

 conduit

0.73

Activations Density 0.191%

phrases related to group membership or identity

terms related to social roles, statuses, and classifications of individuals or groups

No Comments

No Known Activations