INDEX

Explanations

specific patterns of characters, particularly the three characters "fa" followed by numbers or special characters

oai_token-act-pair · gpt-3.5-turbo

words related to facial features or characteristics

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

pin

-0.72

 indo

-0.70

sed

-0.69

tin

-0.68

suit

-0.68

lines

-0.63

tics

-0.63

tor

-0.61

rawdownloadcloneembedreportprint

-0.61

 Norn

-0.60

POSITIVE LOGITS

Ze

1.10

qs

1.07

herty

1.04

isance

1.02

erie

0.99

emia

0.92

Ã§

0.91

resh

0.90

enza

0.90

irst

0.88

Activations Density 0.030%

specific patterns of characters, particularly the three characters "fa" followed by numbers or special characters

words related to facial features or characteristics

No Comments

No Known Activations