INDEX

Explanations

instances of actions or events starting or beginning

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "opened" 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

grade

-0.68

rior

-0.62

orum

-0.62

 Cyborg

-0.60

hov

-0.60

âĨĳ

-0.59

idian

-0.59

ullah

-0.59

iable

-0.59

ivable

-0.58

POSITIVE LOGITS

 Doors

0.92

 doors

0.81

bucks

0.79

 hostilities

0.78

 parentheses

0.78

up

0.77

 wounds

0.74

ococ

0.74

ource

0.73

 curtains

0.71

Activations Density 0.039%

instances of actions or events starting or beginning

occurrences of the word "opened" in various contexts

No Comments

No Known Activations