INDEX

Explanations

the word "refer"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "refer" and its variations

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

alach

-0.78

 whiff

-0.65

 underdog

-0.61

avorite

-0.61

iller

-0.60

onel

-0.58

oppable

-0.58

hov

-0.58

cape

-0.58

olean

-0.57

POSITIVE LOGITS

rers

0.86

irect

0.85

entious

0.84

itatively

0.79

ãĥĥ

0.77

rences

0.76

iously

0.75

 Refer

0.72

oras

0.71

minist

0.71

Activations Density 0.017%

the word "refer"

instances of the word "refer" and its variations

No Comments

No Known Activations