INDEX

Explanations

instances where factual information or knowledge is stated definitively

oai_token-act-pair · gpt-3.5-turbo

assertions and statements of knowledge

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ecycle

-0.63

 upgr

-0.60

phrine

-0.60

attery

-0.60

twitch

-0.58

taking

-0.57

ãĤ¦

-0.57

pex

-0.57

ksh

-0.57

inances

-0.57

POSITIVE LOGITS

 anecd

1.04

 definitively

0.97

 unequivocally

0.92

 that

0.79

 nothing

0.71

 darn

0.69

 unanimously

0.68

 confidently

0.68

 instinctively

0.65

nothing

0.64

Activations Density 0.138%

instances where factual information or knowledge is stated definitively

assertions and statements of knowledge

No Comments

No Known Activations