INDEX

Explanations

proper nouns denoting individuals or groups in positions of leadership or authority

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Lead" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Sabha

-0.80

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.74

irl

-0.72

 exting

-0.69

ALLY

-0.68

 ultras

-0.68

ypes

-0.65

aneously

-0.64

 DISTR

-0.64

ctica

-0.64

POSITIVE LOGITS

 Lead

1.24

Lead

1.19

ership

1.15

lead

0.96

better

0.82

 Developer

0.81

Beat

0.80

 Leading

0.75

found

0.73

 Leader

0.73

Activations Density 0.011%

proper nouns denoting individuals or groups in positions of leadership or authority

occurrences of the word "Lead" in various contexts

No Comments

No Known Activations