INDEX

Explanations

references to hats or actions involving hats

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "hat."

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

MSM

-0.68

 Diagn

-0.68

ateral

-0.67

 Integ

-0.66

CLS

-0.65

Ong

-0.64

theless

-0.63

course

-0.61

 Sacrament

-0.61

houn

-0.61

POSITIVE LOGITS

chet

1.45

chery

1.37

ches

1.08

 brim

1.07

cher

1.05

glers

0.98

eless

0.97

hat

0.96

 hats

0.95

pins

0.89

Activations Density 0.011%

references to hats or actions involving hats

occurrences of the word "hat."

No Comments

No Known Activations