INDEX

Explanations

proper nouns indicating people or places

oai_token-act-pair · gpt-3.5-turbo

the instances of the letter 'o' 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

advertisement

-0.71

maid

-0.66

 heartbeat

-0.59

imaru

-0.58

racuse

-0.58

 noses

-0.57

gd

-0.57

 dstg

-0.57

ron

-0.56

è¦ļéĨĴ

-0.56

POSITIVE LOGITS

ff

0.66

aff

0.63

za

0.63

elt

0.62

efer

0.61

apan

0.60

let

0.60

ort

0.59

iture

0.58

Aff

0.58

Activations Density 0.133%

proper nouns indicating people or places

the instances of the letter 'o' in various contexts

No Comments

No Known Activations