INDEX

Explanations

names or terms related to "Ah"

oai_token-act-pair · gpt-3.5-turbo

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

 fixation

-0.72

DragonMagazine

-0.71

 Colossus

-0.64

etary

-0.62

Closure

-0.60

 spores

-0.60

 sincerity

-0.60

 bottom

-0.59

eering

-0.59

 Daredevil

-0.59

POSITIVE LOGITS

ahah

1.11

oy

1.05

aha

1.02

mad

1.01

renheit

0.92

awk

0.88

ora

0.87

hh

0.87

jong

0.87

ghan

0.86

Activations Density 0.021%

names or terms related to "Ah"

No Comments

No Known Activations