INDEX

Explanations

questions asked rhetorically or seeking confirmation

oai_token-act-pair · gpt-3.5-turbo

rhetorical questions

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

 slam

-0.74

apan

-0.72

 encount

-0.71

 glim

-0.69

ription

-0.68

apers

-0.68

 binge

-0.64

ilogy

-0.63

obos

-0.63

aper

-0.62

POSITIVE LOGITS

 Nope

0.90

Why

0.87

ï¿½ï¿½ï¿½ï¿½

0.84

 Yeah

0.84

 Well

0.83

 Absolutely

0.82

Nah

0.81

 Didn

0.81

 Where

0.79

 Okay

0.79

Activations Density 0.077%

questions asked rhetorically or seeking confirmation

rhetorical questions

No Comments

No Known Activations