INDEX

Explanations

references to political impeachment proceedings

oai_token-act-pair · gpt-3.5-turbo

references to impeachment and related terminology

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

nown

-0.72

Hop

-0.67

hare

-0.67

variable

-0.67

Bee

-0.66

othe

-0.66

WAR

-0.66

VIS

-0.66

ijk

-0.64

 Trader

-0.64

POSITIVE LOGITS

 impeachment

1.29

 impe

1.22

agate

0.86

urated

0.84

 indicted

0.82

 hearings

0.78

cles

0.76

unity

0.76

ptin

0.75

 proceedings

0.75

Activations Density 0.007%

references to political impeachment proceedings

references to impeachment and related terminology

No Comments

No Known Activations