INDEX

Explanations

references to the Democratic Party and its members

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

mentions of political affiliations, specifically the word "Democrat" and similar variants

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

iddles

-0.72

urious

-0.70

ansen

-0.69

utils

-0.69

ysics

-0.66

prototype

-0.65

oken

-0.65

othal

-0.64

ammad

-0.64

tower

-0.64

POSITIVE LOGITS

INTON

0.86

 Party

0.85

 caucuses

0.78

naissance

0.77

 nominee

0.72

 politician

0.72

 senator

0.71

 operative

0.71

hood

0.71

izable

0.70

Activations Density 0.039%

references to the Democratic Party and its members

mentions of political affiliations, specifically the word "Democrat" and similar variants

No Comments

No Known Activations