INDEX

Explanations

names, specifically the name "Les"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Les" 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 @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

INAL

-0.58

REL

-0.58

 clear

-0.57

 sailing

-0.56

 trail

-0.55

 dividing

-0.55

ERA

-0.55

 calculus

-0.54

 converter

-0.54

DPR

-0.54

POSITIVE LOGITS

bians

1.39

nar

1.20

bian

1.19

bos

1.06

lie

1.04

bia

0.98

zek

0.96

ley

0.89

opher

0.81

leys

0.80

Activations Density 0.030%

names, specifically the name "Les"

occurrences of the name "Les" in various contexts

No Comments

No Known Activations