INDEX

Explanations

the word "Lamb" preceded or followed by a specific suffix

oai_token-act-pair · gpt-3.5-turbo

mentions of the name "Lamb" 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 @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

BOOK

-0.84

PLIED

-0.81

 Occupations

-0.71

 Juda

-0.69

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

-0.66

 ignorance

-0.66

pac

-0.65

ãģ¦

-0.64

Downloadha

-0.63

×Ļ×

-0.63

POSITIVE LOGITS

 Lamb

0.86

asted

0.83

inate

0.78

 lamb

0.76

din

0.74

borgh

0.74

aquin

0.73

brace

0.73

adesh

0.72

let

0.72

Activations Density 0.010%

the word "Lamb" preceded or followed by a specific suffix

mentions of the name "Lamb" in various contexts

No Comments

No Known Activations