INDEX

Explanations

terms related to various professions and their interactions, particularly in the context of work and management

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

izoph

-0.76

Reloaded

-0.69

otropic

-0.67

aceous

-0.67

INGTON

-0.66

bush

-0.66

ãĥ´

-0.65

;;;;;;;;;;;;

-0.65

âĶĢâĶĢâĶĢâĶĢ

-0.65

oÄŁ

-0.65

POSITIVE LOGITS

hips

0.95

 alike

0.94

 extraord

0.90

ateurs

0.88

 rejoice

0.84

paces

0.80

hip

0.79

 specializing

0.78

 everywhere

0.75

 keen

0.73

Activations Density 0.174%

terms related to various professions and their interactions, particularly in the context of work and management

No Comments

No Known Activations