INDEX

Explanations

comparisons or contrasts in the form of "more than" statements

oai_token-act-pair · gpt-3.5-turbo

comparative phrases that highlight the complexity or severity of a situation

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

EMENT

-0.81

antage

-0.79

imity

-0.76

uto

-0.76

ance

-0.71

FTWARE

-0.70

urther

-0.70

ulpt

-0.69

bilt

-0.69

autions

-0.68

POSITIVE LOGITS

 usual

0.91

 anything

0.75

 ours

0.73

 ordinary

0.72

 placebo

0.72

 superficial

0.71

average

0.71

 ever

0.70

ĻĤ

0.70

 Watergate

0.67

Activations Density 0.067%

comparisons or contrasts in the form of "more than" statements

comparative phrases that highlight the complexity or severity of a situation

No Comments

No Known Activations