INDEX

Explanations

instances of the word "Nar" or "nar" at different activations

oai_token-act-pair · gpt-3.5-turbo

mentions of the character Naruto

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

 Pixie

-0.84

 Canary

-0.77

 Democr

-0.72

 Extrem

-0.71

xual

-0.69

ORED

-0.68

eous

-0.67

 Giuliani

-0.66

 VIDEOS

-0.65

UCT

-0.65

POSITIVE LOGITS

ration

1.07

rils

0.97

ayan

0.94

uno

0.90

ril

0.90

opa

0.89

vik

0.89

stal

0.89

aku

0.88

vati

0.86

Activations Density 0.026%

instances of the word "Nar" or "nar" at different activations

mentions of the character Naruto

No Comments

No Known Activations