INDEX

Explanations

numerical values with high significance, possibly prices, scores, or rankings

oai_token-act-pair · gpt-3.5-turbo

numerical values and identifiers in a document

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

netflix

-0.77

TextColor

-0.71

rely

-0.70

ques

-0.61

deen

-0.58

rette

-0.58

stood

-0.58

raviolet

-0.57

 Uriel

-0.57

tainment

-0.57

POSITIVE LOGITS

acity

1.12

 Downing

0.92

^{

0.91

tons

0.88

aciously

0.84

acious

0.83

th

0.83

0.82

uther

0.80

0.78

Activations Density 0.101%

numerical values with high significance, possibly prices, scores, or rankings

numerical values and identifiers in a document

No Comments

No Known Activations