INDEX

Explanations

phrases related to actions or events happening in the past or being described in a storytelling manner

oai_token-act-pair · gpt-3.5-turbo

rhetorical questions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

iband

-0.77

 withdraw

-0.68

 administration

-0.66

 payroll

-0.66

 pensions

-0.66

 bloc

-0.66

 Nato

-0.66

achev

-0.66

 administrations

-0.65

endment

-0.65

POSITIVE LOGITS

VICE

0.98

SEE

0.89

His

0.84

Known

0.80

Created

0.80

TED

0.79

ccording

0.79

Specifically

0.77

Interested

0.76

RELATED

0.76

Activations Density 0.627%

phrases related to actions or events happening in the past or being described in a storytelling manner

rhetorical questions

No Comments

No Known Activations