INDEX

Explanations

phrases indicating desire or intention, potentially related to conflict or decision-making

oai_token-act-pair · gpt-3.5-turbo

expressions of reluctance or negative sentiment

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

 SERV

-0.74

 Compan

-0.62

 Windsor

-0.59

enhagen

-0.57

 crosses

-0.57

 successive

-0.57

 contrasted

-0.56

CY

-0.56

 Baldwin

-0.56

 Lancaster

-0.55

POSITIVE LOGITS

0.99

ivably

0.91

be

0.87

¹

0.86

require

0.84

else

0.84

say

0.83

nt

0.83

send

0.83

schild

0.82

Activations Density 0.098%

phrases indicating desire or intention, potentially related to conflict or decision-making

expressions of reluctance or negative sentiment

No Comments

No Known Activations