INDEX

Explanations

mentions of geographical features, specifically cliffs

oai_token-act-pair · gpt-3.5-turbo

instances of the word "cliff" in various contexts

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

Interstitial

-1.04

rity

-0.89

apolis

-0.86

Role

-0.82

alg

-0.79

ufact

-0.79

natureconservancy

-0.78

zona

-0.76

agara

-0.74

algia

-0.71

POSITIVE LOGITS

 cliffs

0.93

 cliff

0.92

 dwellings

0.84

 Dwell

0.79

 stump

0.77

 bluff

0.77

 Cliff

0.75

 ledge

0.75

 Guth

0.74

 swall

0.72

Activations Density 0.016%

mentions of geographical features, specifically cliffs

instances of the word "cliff" in various contexts

No Comments

No Known Activations