INDEX

Explanations

acronyms or abbreviations ending with 'LE'

oai_token-act-pair · gpt-3.5-turbo

references to "LE" or related abbreviations and terms within a specific context

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

ãĤ±

-0.81

ãĥ¼ãĥĨãĤ£

-0.80

ãĤ¢ãĥ«

-0.79

asio

-0.79

raints

-0.78

orship

-0.77

Ì¶

-0.76

Downloadha

-0.76

orial

-0.75

acca

-0.73

POSITIVE LOGITS

LE

1.16

IGH

1.09

VEL

1.04

ASE

0.94

BO

0.91

PE

0.90

lla

0.90

AR

0.89

lean

0.89

LE

0.88

Activations Density 0.005%

acronyms or abbreviations ending with 'LE'

references to "LE" or related abbreviations and terms within a specific context

No Comments

No Known Activations