INDEX

Explanations

phrases indicating a future or current state or situation

oai_token-act-pair · gpt-3.5-turbo

contractions of the word "is" used 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

 Hitch

-0.63

lication

-0.62

idence

-0.61

abad

-0.60

ournals

-0.58

Footnote

-0.58

ience

-0.58

ikers

-0.57

itches

-0.57

ixel

-0.56

POSITIVE LOGITS

 happening

0.94

 transpired

0.86

 meant

0.86

actually

0.81

 happened

0.79

fit

0.76

nant

0.74

 happ

0.73

 termed

0.73

really

0.71

Activations Density 0.124%

phrases indicating a future or current state or situation

contractions of the word "is" used in various contexts

No Comments

No Known Activations