INDEX

Explanations

phrases related to positions of authority

oai_token-act-pair · gpt-3.5-turbo

phrases related to leadership positions and committee roles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 hone

-0.71

 selves

-0.69

sim

-0.69

 plent

-0.66

 distractions

-0.65

 setbacks

-0.65

 thri

-0.64

 depictions

-0.63

 conform

-0.62

 beans

-0.62

POSITIVE LOGITS

 Parliament

0.95

 parliament

0.94

 committee

0.91

 Congress

0.88

 Subcommittee

0.86

 committees

0.85

 Senate

0.84

 Cabinet

0.84

 subcommittee

0.84

 Congressional

0.84

Activations Density 0.128%

phrases related to positions of authority

phrases related to leadership positions and committee roles

No Comments

No Known Activations