INDEX

Explanations

words indicating a progression towards an outcome, result, or decision

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate a sequence of events or outcomes

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

igger

-0.80

ocene

-0.78

 Yard

-0.76

igers

-0.74

adr

-0.73

utenberg

-0.73

raught

-0.73

rieve

-0.72

unity

-0.72

ir

-0.69

POSITIVE LOGITS

 thereafter

0.93

 importantly

0.87

 consequently

0.85

 rely

0.77

 succeeded

0.77

 alike

0.75

 replaced

0.72

 proceeded

0.72

 replaces

0.72

 anticipate

0.69

Activations Density 0.161%

words indicating a progression towards an outcome, result, or decision

phrases that indicate a sequence of events or outcomes

No Comments

No Known Activations