INDEX

Explanations

friendly conversational expressions

oai_token-act-pair · gpt-3.5-turbo

conversational expressions and interactions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Instr

-0.71

Mobil

-0.70

Footnote

-0.65

 Vaugh

-0.62

 Nielsen

-0.60

ãĥ¼ãĥĨ

-0.59

Equ

-0.58

 Barrett

-0.58

 Mobil

-0.58

 Restoration

-0.57

POSITIVE LOGITS

 dont

1.15

 english

1.10

 doesnt

1.08

 didnt

1.04

 alot

1.03

tho

1.03

 americ

1.02

!!!!

0.96

 fuck

0.93

 pics

0.93

Activations Density 0.701%

friendly conversational expressions

conversational expressions and interactions

No Comments

No Known Activations