INDEX

Explanations

terms or phrases related to definitions or interpretations

oai_token-act-pair · gpt-3.5-turbo

instances of the word "definition" and its associated context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

rop

-0.62

imentary

-0.61

 Antar

-0.61

ili

-0.61

 Elect

-0.60

roph

-0.60

itch

-0.59

orld

-0.59

 Pradesh

-0.59

 Observer

-0.58

POSITIVE LOGITS

inition

1.11

 definition

1.07

 definitions

1.06

initions

1.01

definition

0.93

 defines

0.89

 Definition

0.88

witz

0.88

naire

0.83

REDACTED

0.81

Activations Density 0.010%

terms or phrases related to definitions or interpretations

instances of the word "definition" and its associated context

No Comments

No Known Activations