INDEX

Explanations

words with the prefix "en-" or derivatives

oai_token-act-pair · gpt-3.5-turbo

instances of the substring "en" in various contexts

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

INGS

-0.61

 shake

-0.61

ooth

-0.60

TPPStreamerBot

-0.60

 Intermediate

-0.58

THING

-0.57

 distortion

-0.57

 hitch

-0.57

ts

-0.57

 Pyth

-0.56

POSITIVE LOGITS

emies

1.36

abling

1.21

vironments

1.20

closed

1.17

chant

1.17

forcing

1.16

velop

1.15

forcement

1.12

forced

1.10

emy

1.09

Activations Density 0.024%

words with the prefix "en-" or derivatives

instances of the substring "en" in various contexts

No Comments

No Known Activations