INDEX

Explanations

phrases related to the analysis or study of spectrums

oai_token-act-pair · gpt-3.5-turbo

references to "spectroscopy" or related scientific terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

UNCH

-0.65

etheless

-0.62

cooked

-0.62

¶ħ

-0.61

STRUCT

-0.59

 ballpark

-0.59

BY

-0.58

bably

-0.58

UTION

-0.58

æ©Ł

-0.57

POSITIVE LOGITS

rums

1.12

acles

1.01

rosc

1.01

rog

1.00

ograms

0.96

ril

0.92

rils

0.92

uous

0.91

rum

0.88

 spect

0.88

Activations Density 0.015%

phrases related to the analysis or study of spectrums

references to "spectroscopy" or related scientific terms

No Comments

No Known Activations