INDEX

Explanations

references to statements or observations made by someone

oai_token-act-pair · gpt-3.5-turbo

instances where something is being reported or stated

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

heed

-0.81

ciating

-0.73

 waged

-0.67

abase

-0.66

quer

-0.66

thood

-0.65

adesh

-0.65

WAY

-0.65

 collided

-0.63

ravel

-0.62

POSITIVE LOGITS

 notes

0.77

 noted

0.74

 footnote

0.74

unct

0.71

 noting

0.71

 deductions

0.68

edly

0.66

ered

0.66

 Schiff

0.64

 inconsistencies

0.64

Activations Density 0.024%

references to statements or observations made by someone

instances where something is being reported or stated

No Comments

No Known Activations