INDEX

Explanations

mentions of grassy landscapes or objects related to grass

oai_token-act-pair · gpt-3.5-turbo

references to grass

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

osexual

-0.76

umbn

-0.74

icts

-0.66

ukong

-0.65

 Heller

-0.64

 critically

-0.63

andem

-0.63

Fas

-0.63

abol

-0.62

uras

-0.61

POSITIVE LOGITS

roots

1.12

ho

1.01

lands

1.00

nut

0.96

flower

0.96

wood

0.92

 airst

0.92

grass

0.88

woods

0.87

dale

0.87

Activations Density 0.009%

mentions of grassy landscapes or objects related to grass

references to grass

No Comments

No Known Activations