INDEX

Explanations

links to Twitter profiles and tweets

oai_token-act-pair · gpt-3.5-turbo

mentions or references to URLs or social media links

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 afore

-0.69

 Levant

-0.68

 condol

-0.66

 corrections

-0.65

 caring

-0.64

 sanctions

-0.61

 correspondence

-0.61

 deliber

-0.58

 Reck

-0.58

 unsure

-0.57

POSITIVE LOGITS

zx

1.27

zl

1.14

qv

1.14

GY

1.14

zn

1.12

gz

1.11

OX

1.10

zu

1.09

qq

1.09

ZI

1.09

Activations Density 0.050%

links to Twitter profiles and tweets

mentions or references to URLs or social media links

No Comments

No Known Activations