INDEX

Explanations

phrases related to news reporting or quotes by individuals

oai_token-act-pair · gpt-3.5-turbo

sentiments and opinions related to experiences or evaluations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 Whilst

-0.80

 colourful

-0.74

 summar

-0.73

 unsurprisingly

-0.73

 favourite

-0.73

 Cardiff

-0.72

 analysed

-0.72

BBC

-0.69

 organise

-0.68

 criticised

-0.68

POSITIVE LOGITS

 ..."

1.21

.''

1.09

.""

1.09

[/

1.05

----

1.04

-|

1.00

.ï¿½

0.97

,''

0.95

[/

0.95

"""

0.94

Activations Density 1.808%

phrases related to news reporting or quotes by individuals

sentiments and opinions related to experiences or evaluations

No Comments

No Known Activations