INDEX

Explanations

personal pronouns followed by an action or description

oai_token-act-pair · gpt-3.5-turbo

first-person pronouns and assertions of personal experience

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

 unavailable

-0.64

cknow

-0.63

 considerably

-0.62

empt

-0.59

 independently

-0.58

 uncertain

-0.58

 unable

-0.58

knowledge

-0.57

viol

-0.56

Xi

-0.56

POSITIVE LOGITS

 meant

0.90

 envisioned

0.89

 hoped

0.84

 preached

0.80

 stri

0.78

 boils

0.78

 Wanted

0.77

 intended

0.77

 wanted

0.76

 supposed

0.72

Activations Density 0.158%

personal pronouns followed by an action or description

first-person pronouns and assertions of personal experience

No Comments

No Known Activations