INDEX

Explanations

numerical sequences without specific patterns

oai_token-act-pair · gpt-3.5-turbo

numerical identifiers or codes, possibly referencing versions or models

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 sort

-0.77

 waiter

-0.71

 unsupported

-0.71

 accomp

-0.71

 given

-0.70

 whisk

-0.70

 leaflets

-0.70

 prescribed

-0.70

 appealing

-0.68

 numbered

-0.67

POSITIVE LOGITS

EVA

1.14

eva

1.11

hex

1.07

humans

1.07

pac

1.03

earth

1.01

change

1.01

hyde

0.99

ever

0.98

stuff

0.95

Activations Density 0.062%

numerical sequences without specific patterns

numerical identifiers or codes, possibly referencing versions or models

No Comments

No Known Activations