INDEX

Explanations

references to data collection or surveys

oai_token-act-pair · gpt-3.5-turbo

words related to polling or survey results

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Concent

-0.72

 Mons

-0.71

 retreat

-0.70

 sunrise

-0.70

 onward

-0.68

 Ascension

-0.67

 Ruins

-0.67

 Saint

-0.65

 encamp

-0.65

enberg

-0.65

POSITIVE LOGITS

 polled

3.12

poll

1.67

Poll

1.45

 garnered

1.33

OTA

1.18

 garner

1.10

 sque

1.06

ifty

1.02

 polling

0.99

 polls

0.96

Activations Density 0.036%

references to data collection or surveys

words related to polling or survey results

No Comments

No Known Activations