INDEX

Explanations

concrete nouns related to settings or scenarios

oai_token-act-pair · gpt-3.5-turbo

key terms related to roles and responsibilities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

leground

-0.58

ivas

-0.58

anooga

-0.57

act

-0.54

ueller

-0.52

 than

-0.51

erers

-0.50

than

-0.49

OPE

-0.47

 Wilmington

-0.47

POSITIVE LOGITS

*:

0.84

:(

0.70

:[

0.69

âĦ¢:

0.67

:{

0.65

âĺĨ

0.63

pse

0.63

,—

0.62

:-

0.61

Newsletter

0.61

Activations Density 0.932%

concrete nouns related to settings or scenarios

key terms related to roles and responsibilities

No Comments

No Known Activations