INDEX

Explanations

the word "it" followed by a verb

oai_token-act-pair · gpt-3.5-turbo

the negation of actions or statements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Skydragon

-0.70

 arsen

-0.70

izons

-0.69

 Maiden

-0.64

 Houses

-0.63

Ĥ¬

-0.61

 Nile

-0.59

 Presence

-0.58

 Koen

-0.58

izont

-0.58

POSITIVE LOGITS

ogether

0.93

actly

0.92

ember

0.86

Â¨

0.86

withstanding

0.82

prise

0.76

 necessarily

0.76

ymes

0.75

 surpr

0.70

ournament

0.70

Activations Density 0.091%

the word "it" followed by a verb

the negation of actions or statements

No Comments

No Known Activations