INDEX

Explanations

expressions related to making comparisons or evaluating options

oai_token-act-pair · gpt-3.5-turbo

references to measurements, quantities, or numerical values

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Monstrous

-0.56

 entire

-0.54

 outright

-0.53

 stuff

-0.50

 shit

-0.50

 simulac

-0.49

GG

-0.49

 whole

-0.48

 likewise

-0.48

 utterly

-0.47

POSITIVE LOGITS

rin

0.59

rika

0.58

ocrates

0.55

llah

0.54

TPPStreamerBot

0.54

thens

0.54

ãĤ§

0.53

igr

0.53

acement

0.52

Â¯Â¯Â¯Â¯Â¯Â¯Â¯Â¯

0.51

Activations Density 0.763%

expressions related to making comparisons or evaluating options

references to measurements, quantities, or numerical values

No Comments

No Known Activations