INDEX

Explanations

references to the concept of "match" or "matching."

oai_token-act-pair · gpt-3.5-turbo

references to "match" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

Ve

-0.65

 Ways

-0.64

etooth

-0.62

 Lomb

-0.61

ible

-0.61

Asset

-0.58

è¡

-0.58

ennes

-0.58

Balt

-0.57

Coy

-0.57

POSITIVE LOGITS

making

1.38

maker

1.33

stick

1.11

makers

1.10

sticks

1.07

ups

1.02

nings

0.96

up

0.91

day

0.89

match

0.88

Activations Density 0.045%

references to the concept of "match" or "matching."

references to "match" in various contexts

No Comments

No Known Activations