INDEX

Explanations

mentions of specific phrases containing the word "enz" in them

oai_token-act-pair · gpt-3.5-turbo

mentions of specific locations and associated names

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

GROUND

-0.90

FUL

-0.75

 Pend

-0.73

bable

-0.73

 Courage

-0.69

TAIN

-0.68

WOOD

-0.68

ful

-0.65

READ

-0.65

 Scots

-0.63

POSITIVE LOGITS

irez

1.04

enz

1.04

ymes

1.04

arella

1.03

etta

1.01

hou

0.96

hao

0.96

anz

0.94

akura

0.88

enger

0.86

Activations Density 0.020%

mentions of specific phrases containing the word "enz" in them

mentions of specific locations and associated names

No Comments

No Known Activations