INDEX

Explanations

names that contain the substring "der"

oai_token-act-pair · gpt-3.5-turbo

instances of the suffix "der" indicating a specific pattern in words

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ainted

-0.64

INTER

-0.60

Memory

-0.60

ELS

-0.59

 Butcher

-0.58

Sri

-0.58

pta

-0.56

FIX

-0.56

ERC

-0.56

WS

-0.55

POSITIVE LOGITS

dash

1.01

minster

0.85

ricks

0.84

iving

0.83

rors

0.82

hoe

0.82

netic

0.81

vous

0.80

isively

0.79

theless

0.78

Activations Density 0.012%

names that contain the substring "der"

instances of the suffix "der" indicating a specific pattern in words

No Comments

No Known Activations