INDEX

Explanations

accurate descriptions of passing actions, possibly in sports or other contexts

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "passes"

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

birds

-0.67

ASP

-0.66

 Providence

-0.65

 Stevens

-0.63

bird

-0.63

aiman

-0.63

 prin

-0.60

antha

-0.59

 stressed

-0.59

rob

-0.58

POSITIVE LOGITS

ãģĨ

1.00

 passes

0.95

wd

0.92

words

0.87

ologne

0.84

 Passed

0.84

 Passing

0.84

engers

0.80

phrase

0.80

itters

0.73

Activations Density 0.010%

accurate descriptions of passing actions, possibly in sports or other contexts

occurrences of the word "passes"

No Comments

No Known Activations