INDEX

Explanations

phrases that describe actions and situations involving oneself in various contexts

oai_token-act-pair · gpt-3.5-turbo

prepositions indicating position or situation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

rar

-0.70

heny

-0.68

 Feature

-0.67

liament

-0.67

 sqor

-0.66

ï¸ı

-0.66

gap

-0.65

ctory

-0.64

illary

-0.62

payer

-0.61

POSITIVE LOGITS

isner

0.65

 nomine

0.65

 pant

0.63

swer

0.63

 toile

0.60

peror

0.59

 hunted

0.59

 Valiant

0.59

 Diet

0.58

 selves

0.58

Activations Density 0.191%

phrases that describe actions and situations involving oneself in various contexts

prepositions indicating position or situation

No Comments

No Known Activations