INDEX

Explanations

phrases related to tasks or instructions

oai_token-act-pair · gpt-3.5-turbo

phrases and words that emphasize personal agency or self-directed actions

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

inguished

-0.77

Applications

-0.73

icularly

-0.73

ertility

-0.71

Ĥª

-0.70

Enhanced

-0.68

perties

-0.67

edIn

-0.65

letal

-0.65

ustomed

-0.64

POSITIVE LOGITS

 clause

1.03

 button

1.02

 factor

0.86

"-

0.83

 drawer

0.80

'-

0.78

 clauses

0.77

 bandwagon

0.77

 option

0.76

Activations Density 0.177%

phrases related to tasks or instructions

phrases and words that emphasize personal agency or self-directed actions

No Comments

No Known Activations