INDEX

Explanations

descriptions of research studies involving large sample sizes

oai_token-act-pair · gpt-3.5-turbo

references to statistical data and surveys involving families or groups

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

eln

-0.66

Fax

-0.65

#$

-0.65

Nevertheless

-0.63

Tro

-0.62

Nonetheless

-0.62

Nib

-0.62

EXT

-0.61

HCR

-0.60

}}}

-0.60

POSITIVE LOGITS

 sampled

1.20

 conducted

1.14

 surveyed

1.06

 spanning

1.05

 analyzed

1.01

 collected

1.00

 analysed

0.97

 undertaken

0.96

 comparing

0.95

 across

0.94

Activations Density 0.348%

descriptions of research studies involving large sample sizes

references to statistical data and surveys involving families or groups

No Comments

No Known Activations