INDEX

Explanations

references to tracking or following something

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "trace" or "tracing."

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

otos

-0.82

comes

-0.71

icum

-0.70

incial

-0.69

ows

-0.68

itch

-0.66

annis

-0.65

IDES

-0.65

iterranean

-0.63

skill

-0.63

POSITIVE LOGITS

 traces

1.03

 tracing

1.00

 trace

0.98

 traced

0.98

trace

0.82

 Trace

0.79

 Armstrong

0.76

ability

0.76

 chronological

0.73

 bracelet

0.69

Activations Density 0.019%

references to tracking or following something

references to the concept of "trace" or "tracing."

No Comments

No Known Activations